[metapost] Trying to use MT1 to make outline fonts... (again)

Shriramana Sharma samjnaa at gmail.com
Tue Jun 19 09:14:18 CEST 2012

Begging the patience of those who have already replied to my similar
questions previously.

I have taken the time since those previous mails to read through
various documents (mostly TugBoat articles) available on this subject
on the web, and to procure a copy of the MetaFontBook for proper
reading. (I'm into chapter 12 Boxes right now after skipping the
somewhat dry chapters on how MF parses tokens and what MF's language
grammar is.)

I also read http://existentialtype.net/2009/04/18/testing-mftrace/
which says mftrace is still fine. The outlines produced by mftrace are
not as per font conventions though [regarding extrema and the like].

I have also in the meanwhile produced some glyphs for a Unicode
proposal I'm working on using Inkscape, by constructing a path, apply
a stroke width and telling Inkscape to convert strokes to paths,
exporting to PDF and importing in High-Logic Font Creator, scaling it
to required size, applying proper bearings etc.

However I'm not satisfied with the mathematical beauty of my glyphs
(due to some limitations of Inkscape, great software though it is) and
still want to try the Meta* approach to font making.

A friend very clearly recommended that MetaType1 is what I want to
create outline fonts using the MetaFont approach. Frankly to me also
this is what seems the most appropriate, especially seeing Piska's
article on creating a Unicode cuneiform font using it
(www.tug.org/TUGboat/tb29-3/tb93piska.pdf) which is very close to my
intention of creating fonts for epigraphical scripts of India.

I do have some questions about MT1 which I request to be cleared and
which I was unable (sorry to say) to clear myself despite reading all
those TugBoat articles. (Many of them seem to be targeted at
experienced TeX users which I am not. And there is a dearth of
detailed newbie-targeted explanation of how to use MT1. Also, though I
have the MFBook now, I'm not sure how much of it applies to MT1 too...

The following are my questions:

Piska's older article (www-hep2.fzu.cz/~piska/TUG2004/piskatb2.pdf)
says about MT1:

on pp 5, 6:

The second problem is that in MetaType1 (I used MetaType1 version 0.40
of 2003) a regular pen stroking algorithm is not available, only a
simplified method of connecting the points ‘parallel’ to the nodes on
the path ... I hope that this problem could be solved in a future
release, at least for pen stroking with a circular pen.

Even more serious is a situation with the rotated elliptic pen ...
Unfortunately, in this case the results of the envelope approximation
are not correct and cannot be used.

on p 10:

* bad pen stroking algorithm; in particular, results for complicated
fonts using rotated elliptical pens are unusable

I would like to know:

1) Has this "regular pen stroking algorithm" since the article been
added to MT1?

2) Are rotated elliptical pens usable?

3) Even if not, are non-rotated elliptical pens usable?

4) Even if not, at least are circular pens usable?

If rotated elliptical pens are usable, great! It would help me create
more stylistic fonts. But if not, I would like to at least create
sedate fonts with a "soft" round "felt tip" pen. (After all the Indian
kings/others of yore who employed those scripts are not going to
commission me to create more and more beautiful typefaces!)

5) The MFBook tells me to do pickup pencircle xscaled yscaled rotated
and do "penstroke". It is not very clear to me whether I can do
penstrokes in MT1 or I have to do outlines only and fill them. (Means
the computer doesn't take care of converting the strokes to paths.)
Can anyone please clarify?

I'm very sorry if my questions seem childish or repetitive (despite
the answer being "out there" somewhere) but I'm really at the end of
my tether here and want to get on with my epigraphic encoding project
as soon as I can get help from some kind soul(s) here.

Please! Help!

Shriramana Sharma

More information about the metapost mailing list